Do you know what ServiceNow application nodes are?
Most of us just refer to them as "nodes", but even within that, did you know that there are actually different types of nodes?
Every ServiceNow instance has at least 2 application nodes and in some rare cases a specific instance/database could have up to 37 application nodes on it, which is roughly the max amount an instance could support.
An application node consists of 3 pieces:
- Java servlet (known as Glide) and runs the core platform code
- Servlet container supporting the Java servlet mentioned above
- and a Proprietary tool which links the Java servlet and Servlet container together
More specifically, within each application node there's a pool of threads which execute the work and a JDBC driver and a pool of database connections which allow interaction with the database tier that's underlying. Each application node can technically be deployed as a different type. Commonly, all application nodes are deployed as "Generic" which does all the processing of various workloads such as user generated, integrations, scheduled jobs on the one node with one pool however, there is another node type known as: UI/Worker (which essentially splits the duty to two different pools).
- UI: processes only user and integration/scheduled job transactions that are pinned to that specific node - and points to a pool of UI nodes
- Worker: processes only scheduled jobs and integration transactions if they're configured accordingly - and points to a pool of worker nodes (you may see the URL such as: having "instancenameworker" then service-now in the URL)
When this type of node is deployed, work is split between those two node pools. When needed, ServiceNow would assist with deploying targeted UI/Worker application node(s), otherwise, Generic is used.
When would you want to use a UI/Worker type over Generic? This would be dependent upon the situation such as scheduled jobs causing memory contention, end users are exposed to performance degradation anytime a job executes, or the instance is unstable during the job. ServiceNow won't immediately jump to switching a generic node to a UI/Worker as it increases administrative overhead and reduces resources available to the instance as there's less total threads available following the split, so they'll want to work with you to find out your issues.
You can find out which type a node is by looking at the sys_cluster_state table and looking for a status of online and within the scheduler column (or look at the Node Type column as well) look for: "any" = generic, "specified" = UI/Worker.
It's always recommended to speak to your ServiceNow Account Team for more information related to nodes.
If you enjoy ServiceNow content like this, be sure to follow me on LinkedIn and subscribe to my ServiceNow focused YouTube channel: Allenovation!
Is it upgrading?
I only have these types available
